A Signal Processing Approach for Eucaryotic Gene Identification

Abstract

Signal processing has a great role in innovative developments in technology. Its techniques have been applied mostly in every field of science and engineering. In the field of bioinformatics it has to play an important role in the study of biomedical applications. Accurate prediction of protein coding regions (Exons) from genomic sequences is an increasing demand for bioinformatics research. Many progresses made in the identification of protein coding regions during the last few decades. But the performances of the identification methods still required to be improved. This paper deals with the identification of protein-coding regions of the DNA sequence mainly focus on analysis of the gene introns. Applications of signal processing tools like spectral analysis, digital filtering of DNA sequences are explored. It has been tried to develop a new method to predict protein coding regions based on the fact that most of exon sequences have a 3-base periodicity. The period-3 property found in exons helps signal processing based time-domain and frequency domain methods to predict these regions efficiently. Also, an efficient technique has been developed for the identification of protein coding region based on the period-3 behavior of codon sequences. It is based on time domain periodogram approach. Here it has been identified the protein coding regions, wherein we reduced the background noise significantly and improve the identification efficiency. In addition to this also comparison is done between time domain periodogram and the existing frequency based techniques. Simulation results obtained are shown the effectiveness of the proposed methods. This proves that the DSP techniques have important applications in obtaining useful information from these gene sequences.